To test the hypothesis that withdrawal of neurohormonal activation following conversion of atrial fibrillation exaggerates the QT prolonging effects of class III antiarrhythmics.